Shenzhen’s initiative for more efficient coordination and allocation of AI computing resources reflects the urgency in China to accelerate development in this field amid US trade sanctions, which have blocked mainland firms’ access to most advanced semiconductors and chip-manufacturing equipment.
Chinese institutions have so far launched at least 79 LLMs with over 1 billion parameters, a measure of the size and complexity of a model. LLMs, the technology behind ChatGPT, are deep-learning algorithms that can recognise, summarise, translate, predict and generate text and other content based on knowledge gained from massive data sets.
Shenzhen’s action plan will also involve building an AI supercomputing network for the Greater Bay Area, bringing together intelligent computing power of governments, enterprises and research institutions in this megalopolis that consists of Hong Kong, Macau and nine cities in Guangdong.
The southern tech hub pledged to support research and development on areas such as AI foundation models and chips to help achieve breakthroughs.
The city will also actively promote and encourage the application of AI technology in a range of sectors including public services, city governance and various industries. Examples include deploying robots for street sweeping and inspection, and adopting AI for image recognition and video analysis to ensure safety in fire control and building construction, according to the action plan.
China has put forward a series of policies, from the national to city level, to ratchet up the country’s efforts in AI, following the success of US start-up OpenAI’s ChatGPT service, which is currently not available in the world’s second-biggest economy.
The mainland plans to establish a raft of regional AI “highlands” across the country and related tech platforms, according to Wang Zhigang, China’s Minister of Science and Technology, at the opening of the 7th World Intelligence Conference in the northern metropolis of Tianjin earlier this month.
